Barley yield and malt quality affected by fall and spring planting under rainfed conditions

Plant materials

Field trials were carried out in the experimental fields of Tokat Agricultural Application and Research Center of Tokat Gaziosmanpasa University (40°33′N, 36°47′ E, 539 m a.s.l) in 2015/16 and 2016/17 in Tokat province of Turkey (Fig. 1) under rainfed conditions. Seven winter, eight facultative and a spring malting barley cultivar Harrington were examined (Table 1).

Field trials

Experimental soils had clayed loam texture, slight levels of salt (0.041 and 0.042%), mild alkaline reaction (pH: 7.77 and 7.80), moderate amounts of lime (10.2 and 7.7%), high and low amounts of P₂O₅ (111 and 58 kg/ha), high levels of K₂O (1108 and 1015 kg/ha) and low levels of organic matter (1.45 and 1.22%) in 2015/16 and 2016/17 years, respectively (Table 2) (Soil analyses were carried out by soil laboratory of the Middle Black Sea Transitional Zone Agricultural Research Institute, Tokat-Turkey). The average temperature and monthly total precipitation of November-June period for long term (48 years) and experimental years are given in Fig. 2. The average long-term precipitation was 359 mm, and the precipitation in 2016 (315 mm) was slightly less than the long term while the precipitation in 2017 (229 mm) was considerably less than the long-term average. The average precipitation of the spring-planting trials was 195 mm in long term (March–June). In 2016, the precipitation (211 mm) was similar to that in long term while the precipitation in 2017 (139 mm) was considerably less. According to these data, it would not be wrong to consider 2016 as a high rainfall environment and 2017 as a low rainfall environment. The trials were carried out in randomized complete block design with four replications. Each plot consisted of five rows of 4 m long. Row spacing was 20 cm. The seeding rate was 500 plants m². Fertilizers were applied to plots as 80 kg/ha P₂O₅ (triple super phosphate) and 80 kg/ha N (Ammonium nitrate).

Measured traits

Time to heading was the period from sowing to first awn occurrence in 50% of the plants in the plot (Saygili & Kandemir, 2021). Plant height was the distance between the ground and the top spikelet in spike except for awn in 15 random plants, and lodging is a visual estimation of the plants lodged in the plot (Kandemir et al., 2000). Maturity was the period from sowing to the time when all leaves turned to yellow (Kandemir et al., 2022). Number of spikes per square meter was calculated by dividing the grain yield (g/m²) by number of grains per spikes and thousand-seed weight (Saygili & Kandemir, 2021). Number of grains per spikes, grain yield, test weight and thousand-seed weight were determined according to Aisawi et al. (2015). Spike length and number of grains per spikes were determined in 30 randomly selected spikes from the plots. Grain yield was determined by converting the grain product obtained from 4 m² plots to t/ha. Thousand-seed weight was determined by counting and weighing 400 random grains. Test weight was determined by weighing the free-falling grains into an exact volume of 250 ml and converting them to hectoliters. Moisture content of grain was determined by drying for 48 h at 75  °C. Grain yield, test weight and thousand-seed weight were calculated on 12% moisture content basis (Kandemir et al., 2022).

Malt production was performed as described by Saygili et al. (2021). Grain was immersed in two cycles soaking, nine hours each, in water and 16 h of draining. Grains were germinated at 14 °C for 110 h. Moisture level was maintained at 45% of the starting grain weight. When the plumula grew to 75% of the grain length, germination was ended. Kilning was performed in consecutive steps of eight hours at 60 °C, six hours at 70 °C and five hours at 80 °C. The rootlets were manually removed. Malt extract was determined using the method described by Fox & Henry (1993) with some modifications. A total of 12 ml (approach by weighing) of distilled water at 65 °C was added to 3 g of ground malt passed through a 0.5 mm sieve. The mixture was incubated in a 65 °C water bath for 60 min and was centrifuged for 5 min at 3000 g. The supernatant was measured with a refractometer. Malt extract was determined according to the formula: ${\displaystyle \text{Malt extract}\left(\text{\%}\right)=\frac{\left(\left[\text{sample}\left(3\mathrm{g}\right)+\text{amount of water}\right]\ast \text{refractometer measurement}\ast 100\right)}{\text{\%}\text{dry matter}}.}$

Diastatic power was measured using the method described by Fox et al. (1999). A total of 10 ml of extraction solution (0.5% NaCl) was added to 1 g of malt and vortexed. The mixture was incubated in a water bath at 25 °C for 30 min. It was centrifuged for 5 min at 2,000 g. A total of 5 ml of buffered starch solution (2% starch, 2 mM glacial acetic acid and 0.05 M sodium acetate, pH: 4.6) was pre-incubated at 25 °C for 5 min. The enzyme extraction supernatant (100 µl) was added to the buffered starch solution and incubated exactly for 10 min at 25 °C and then 1 ml of 0.5 M sodium hydroxide was added to stop the reaction. Five ml of PAHBAH (p-hydroxy benzoic acid hydrazide (5 g/l) dissolved in alkaline diluent, 0.05 M trisodium citrate, 0.01 M calcium chloride, 0.5 M sodium hydroxide) was pre-incubated in a boiling water bath for at least five min. One hundred µl of hydrolysed sample was added to incubated PAHBAH solution and was kept in boiling water for exactly four min. Content is rapidly cooled to room temperature in slurry ice. It was diluted 10-fold and measured at absorbance at 415 nm. Maltose equivalents are determined according to maltose standards curve absorbances (0, 1, 2, 3, and 4 mg/l) at 415 nm. Diastatic power was determined as Windisch-Kolbach unit (°WK) with the following formula. °WK= (87.5 × maltose equivalent) − 16.

Alpha amylase activity was analyzed using a commercial kit (Megazyme International Ireland Limited, Product code: K-CERA, Wicklow, Ireland) based on manufacturer’s instructions. Alpha amylase activity was expressed as Ceralpha units (CU). CU could be converted to dextrinizing unit ASBC method (DU) and AACC method (SKB) with the following formulas: DU = 0.23 × Ceralpha units + 0.61 and SKB units = 0.42 × Ceralpha units − 0.34, respectively.

Statistical analyses

Since variances of years and planting date were not homogeneous based on Barlett’s homogeneity test (p > 0.05), variance analyses of years and planting date were performed separately (Saygili et al., 2021; Bilate Daemo et al., 2023) using JMP Pro 14 software (SAS Institute Inc., Cary, NC, USA). Differences between means were grouped by Tukey multiple comparison test (p < 0.05). GGE biplot and stability analysis were conducted using the GEA-R software (Pacheco et al., 2015) according to Kandemir et al. (2022) and Kandemir (2022) respectively. To compare cultivars for two traits, a scatter plot graphic was drawn using Minitab (ver.17). PCA-Biplot was used for trait-based scaling of cultivars separately for fall and spring trials.

Time to heading

A total of four trials, two fall-planted and two spring-planted, were conducted to determine the performance of the malting barley cultivars. Since cultivar Cumra-2001 produced very few spikes in spring-planted trials, its data in the spring-planted trials were not included in the analyses. The heading time of the cultivars ranged from 163.5 to 178.3 days and from 69.9 to 88.8 days in fall- and spring-planted trials, respectively (Table 3). In fall-planted trials, Cumra-2001 (184.7 and 166.0 days in 2016 and 2017, respectively) and Aydanhanim (187.3 and 171.0 days in 2016 and 2017, respectively) were the latest in heading, while Zeynelaga (173.7 and 155.5 in 2016 and 2017, respectively) and Bolayir (174.3 and 157.5 days in 2016 and 2017, respectively) were the earliest (p < 0.05). In spring-planted trials, Aydanhanim (82.7 and 104.8 days in 2016 and 2017, respectively) reached to heading in significantly longer periods than other cultivars. Unlike the fall-planted trials, Sladoran had late heading in spring-planted trials (78.6 and 106.5 days in 2016 and 2017, respectively). Although Durusu, Yildiz, Bolayir and Sladoran reached to heading early in fall-planted trials, their headings were late in spring-planted trials.

Time to maturity

Time to maturity varied between 199.7 and 227.4 days in fall-planted trials and between 111.9 and 113.1 days in spring-planted ones in 2016 and 2017, respectively (Table 3). In both fall-planted trials, Aydanhanim (210.8 and 231.7 days), Cumra-2001 (211.5 and 232.3 days) and Durusu (210.5 and 230.0 days) matured latest, while Catalhoyuk (192.5 and 223.0 days), Efes-98 (191.5 and 223.0 days), Erciyes (192.5 and 225.0 days) and Ozdemir-05 (193.3 and 225.0 days) matured earliest (p < 0.05 days). In spring-planted trials, Aydanhanim matured late in both years. While Durusu, Harrington, Sladoran and Yildiz were among the cultivars that matured late in the 2016 trial, in 2017 they matured somewhat early. In spring-planted trials, Catalhoyuk was earliest in 2016S (105.7 days), while in 2017S Basgul (103.5 days), Efes-98 (105.5 days) and Zeynelaga (103.5 days) matured relatively earlier than other cultivars. Sladoran, Yildiz and Zeynelaga matured consistently late in 2016S trial and early in 2017S trial.

Plant height

In fall-planted trials, there was a difference of 24 cm in average plant height between 2016 and 2017 trials (84.9 and 109.2 cm, respectively) while in spring-planted trials the difference was only 4.2 cm. In fall-planted trials, Aydanhanim, Basgul, Cumra-2001 and Efes-98 had consistently taller plants (Table 3, p < 0.05). Catalhoyuk, Erciyes, Ince-04 and Tokak 157/37 had taller plants in 2016F trial but relatively moderate plant heights in 2017F trial. Sladoran had consistently low plant heights (68.0 and 88.1 cm) in the two fall-planted trials. In 2016S, Aydanhanim had the tallest plants (98.8 cm), while in 2017S, Basgul, Catalhoyuk, Efes-98, Ince-04 and Tokak 157/37 had taller plants (88.3–95.6 cm). Sladoran and Zeynelaga had shorter plants in the two spring-planted trials.

Lodging

In fall-planted trials, lodging was 46.5 and 67.7% in 2016 and 2017, while in spring-planted it was 47.2 and 8.5% in 2016 and 2017, respectively (Table 4). In fall- and spring-planted trials, Basgul, Catalhoyuk, Efes-98, Erciyes, Kalayci-97, Ozdemir-05 and Tokak 157/37 had consistently higher lodging values (p < 0.05). On the other hand, Bolayir, Cumra-2001, Durusu, Sladoran, Yildiz and Zeynelaga had the lowest lodging values.

Number of grains per spike

Number of grains per spike was similar in all trials except for 2017F. The highest numbers of grains in fall-planted trials were obtained from Aydanhanim (30.3 and 30.5) and Cumra-2001 (29.5 and 30.8) (Table 4, p < 0.05). All cultivars other than Basgul, Bolayir, Durusu, Harrington, Sladoran and Yildiz had the lowest number of grains per spike in fall-planted trials. In spring-planted trials, the highest number of grains per spike was obtained from Aydanhanim (29.5 and 28.9) and Harrington (27.4 and 28.0) in the two years. All cultivars except for Yildiz (27.3 and 24.7) had less seeds per spike (20.5–25.4) in spring-planted trials.

Number of spikes per square meter

The number of spikes per square meter ranged from 421.4 to 483.3 in fall-planted trials and from 302.3 to 433.8 in spring-planted trials (Table 4). In fall-planted trials, the highest number of spikes per square meter was obtained from Harrington as 537.7 and 475.9, from Ince-04 as 532.6 and 499.9 and from Zeynelaga as 547.5 and 470.5 in 2016 and 2017, respectively (p < 0.05). In addition, Bolayir and Ozdemir-05 produced more spikes per square meter in 2016F trial and Durusu, Kalayci-97 and Sladoran produced more spikes in 2017F trial. Cumra-2001 had the lowest number of spikes per square meter in both years. Zeynelaga produced more spikes per area in the two spring-planted trials, while Erciyes, Ozdemir-05 and Sladoran produced more spikes than other cultivars in 2017S trial only. Yildiz had the lowest number of spikes per square meter in the two spring-planted trials, while Aydanhanim and Efes-98 in the 2016S trial only.

Thousand-seed weight

The thousand-seed weight of cultivars varied between 39.7 and 50.6 g in fall-planted trials and between 44.7 and 49.0 in spring-planted trials (Table 5). Durusu had higher thousand-seed weights in the two fall-planted trials (57.7 and 45.1 g in 2016 and 2017, respectively), while Tokak 157/37 had higher values in 2016F trial (59.3 g), and Zeynelaga (46.1 g) and Yildiz (44.6 g) had higher values in 2017F trial (p < 0.05). In the two spring-planted trials, Durusu (52.0 and 48.5 g), Ince-04 (52.0 and 50.7 g) and Tokak 157/3 (53.6 and 50.4 g) had larger seeds while Basgul (50.4 g) and Zeynelaga (49.4 g) produced higher thousand-seed weights in the 2017S trial.

Test weights

In fall-planted trials, Aydanhanim (66.7 kg), Cumra-2001 (67.5 kg), Ince-04 (66.7 kg), Yildiz (66.6 kg) and Zeynelaga (67.3 kg) had higher test weights (Table 5, p < 0.05). Ince-04 had higher test weights in the two spring-planted trials (66.1 and 65.3 kg in 2016 and 2017, respectively). The highest test weight was obtained from Zeynelaga(66.9 kg) in the 2017S trial.

Grain yields

Grain yields varied between 4.38 and 5.71 t/ha in fall-planted trials and between 3.12 and 4.89 t/ha in spring-planted trials (Table 5). Aydanhanim (6.88 t/ha) and Ince-04 (6.42 t/ha) had higher grain yields in 2016F trial while Durusu (5.79 t/ha) and Yildiz (5.33 t/ha) in 2017F trial (p < 0.05). Durusu (6.07−3.23 t/ha), Ince-04(6.19−3.42 t/ha) and Zeynelaga (5.75−3.58 t/ha) had higher grain yields in the two spring-planted trials. In the GGE biplot of grain yield, where PC1 was 67.05 and PC2 was 22.8, Durusu, Ince-04, Erciyes, Tokak 157/37, Efes-98 and Aydanhanim were genotypes located at the extremes (Fig. 3A). Durusu and Ince-04 were the best cultivars in fall-planted trials. There was no environment where other cultivars were best in terms of genotype and genotype x environment interaction. In Fig. 3B, where stability and average yields were compared, the line shown by a circled arrow is average-environment coordination. The arrow indicates the direction in which the means were highest (Frutos, Galindo & Leiva, 2014). The cultivars with the highest yield averages were Durusu, Ince-04, Zeynelaga, Harrington, Aydanhanim, Bolayir and Yildiz. Accordingly, the cultivars closest to the average yield line are the most stable ones (Frutos, Galindo & Leiva, 2014). The most stable cultivars with above the average grain yields were the Zeynelaga, Ince-04 and Durusu. Aydanhanim, Yildiz and Bolayir, on the other hand, were unstable.

Alpha amylase activity

Alpha amylase activities ranged from 77.9 to 81.4 CU/g in fall-planted trials and from 80.8 to 100.9 CU/g in spring-planted trials (Table 6). Yildiz had the highest Alpha amylase activity in the two fall-planted trials (153.8 and 212.2 CU/g in 2016 and 2017, respectively), and Sladoran (133.7 CU/g) had the highest alpha amylase activity in 2017F trial (p < 0.05). In the two spring-planted trials, Harrington(190.1–201.1 CU/g) had the highest alpha amylase values.

Diastatic power

Diastatic power ranged from 194.5 to 331.1° WK in fall-planted trials and from 129.0 to 259.8° WK in spring-planted trials (Table 6). Harrington (611.1° WK) had the highest diastatic power in the 2016F trial, while in the 2017F trial, Bolayir (372.1° WK) and Durusu (320.4° WK) had the highest values (p < 0.05). In spring-planted trials, Bolayir (452.3° WK) had the highest diastatic power in the 2016S trial and Zeynelaga (301.3° WK) in 2017S trial.

Malt extract

In fall-planted trials, average malt extract was 78.0% in 2016 and 77.0% in 2017, which were 76.9 and 73.9 in spring-planted trials in 2016 and 2017, respectively. In fall-planted trials, Durusu, Harrington and Ince-04 had consistently higher malt extract (Table 6, p < 0.05). However, Yildiz had higher malt extract percentage in 2016F trial and Aydanhanim and Bolayir had higher malt extract percentages in the 2017F trial. In spring-planted trials, Harrington (80.7 and 81.9%) and Ince-04 (79.1 and 81.7%) had higher malt extract percentages in both years, while Yildiz had a higher malt extract percentage only in the 2016S trial. In which-won-where view of the GGE biplot, Harrington, Yildiz, Erciyes, Kalayci-97 and Efes-98 were located at the vertices of the polygon, which indicated that these cultivars had the best or poorest performance in an environment or environments (Fig. 4A). Harrington and Yildiz were the best cultivars in terms of malt extract percentages. Ince-04, Durusu, Bolayir, Sladoran, Zeynelaga and Aydanhanim also performed well in all trials. In Fig. 4B, the cultivars with the highest malt extract percentages in decreasing order were Harrington, Ince-04, Yildiz, Durusu, Bolayir, Zeynelaga, Aydanhanim and Sladoran. Among the cultivars with high malt extract percentages, Harrington, Ince-04 and Durusu were relatively stable ones, while Yildiz and Aydanhanim were unstable ones.

Evaluation of all traits

A principal component analysis was carried out in order to evaluate all traits. The first two principal components (PC) accounted for 74.23% of the total variation in fall trials and 77.85% in spring trials (Fig. 5). The most predominant characters in fall trials were lodging, grain yield and malt extract on PC1, and lodging, diastatic power and plant height in spring trials. Pattern of lodging in both planting times was associated with less malt quality. Malt extract, alpha amylase, and diastatic power were closely related in both planting time. In the winter trials Yildiz, Ince-04 and Durusu, and in the spring trials Harrington and Durusu had a trend in the same direction with malt quality criteria such as malt extract, diastatic power and alpha amylase activities. Zeynelağa had the same pattern with the number of ears per square meter in both trials. In fall trials, Durusu and Yildiz showed a similar pattern with grain yield, and Ince-04 in spring trials.